A long-term implantable tissue glucose sensor that can monitor blood glucose concentration is being developed. The sensor is intended to address the most important and difficult challenge for any glucose sensor, namely: reliable detection of mild hypoglycemia in time that action can be taken to prevent hypoglycemia from becoming severe. Further, hypoglycemia may occur during various physiologic states that can interfere with its detection, and it is essential to avoid false positives. There is evidence that temperature, lactate, heart rate and breathing rate may be physiologically coupled with hypoglycemia, although the degree of coupling of these parameters has not been documented under daily life conditions. The glucose sensor array has sensors for these parameters. The proposal is to monitor these variables in parallel to glucose during hypoglycemic transients to determine if they are coupled with hypoglycemia and can confirm detection of mild hypoglycemia by the glucose sensor, thereby reducing the chance of false positives and negatives. Sensor arrays with respective wireless telemetry units will be implanted in pigs and the glycemic response will be studied extensively under a variety of conditions including exercise, rest, and sleep, fasting and feeding. The role of tissue encapsulation and its effects on sensor sensitivity to glucose will be determined as a basis for recalibration. For data analysis, the parallel recordings will be fed into a trainable, pattern recognition algorithm to create a reliable mild hypoglycemia detection system. The data will also be inputted into autoregressive moving average (ARMA) models to determine if hypoglycemia can be predicted, in addition to being detected in real time. Questions about the optimal tissue implant site and whether tissue glucose can lead as well as lag blood glucose will be addressed. A long-term stable sensor for lactate will be developed and implemented in hypoglycemia detection. There is a crucial need to address the growing problem of diabetes. This project will further validate an implantable glucose sensor that will provide a basis for reliable detection of hypoglycemia and automatic insulin administration. [unreadable] [unreadable] [unreadable]